rus
Issue #7/2022
V. A. Zheltikov, D. D. Platonov, S. Khydyrova, D. D. Vasilev, K. M. Moiseev
Review of Superconducting Microstrip Single-photon Detectors
Review of Superconducting Microstrip Single-photon Detectors
DOI: 10.22184/1993-7296.FRos.2022.16.7.528.537
This paper reviews the state-of-the-art of superconducting microstrip single-photon detectors and provides their comparison with the nanostrip detectors. The article describes the actual detection models required for understanding the SMSPD operating principles. We have analyzed the manufacturing materials of ultrathin films designed for SMSPD demonstrating the prospects for the use of X-ray amorphous materials with α-Mn structure. As a result of methods study to improve the detector specifications, we have provided the following recommendations: to reduce DCR and increase SDE, CR and active area, it is necessary to use the SMSPDs based on the X-ray amorphous materials with a low diffusion coefficient, a topology with a high filling factor and the strip spreading at the turn based on a DBR structure.
This paper reviews the state-of-the-art of superconducting microstrip single-photon detectors and provides their comparison with the nanostrip detectors. The article describes the actual detection models required for understanding the SMSPD operating principles. We have analyzed the manufacturing materials of ultrathin films designed for SMSPD demonstrating the prospects for the use of X-ray amorphous materials with α-Mn structure. As a result of methods study to improve the detector specifications, we have provided the following recommendations: to reduce DCR and increase SDE, CR and active area, it is necessary to use the SMSPDs based on the X-ray amorphous materials with a low diffusion coefficient, a topology with a high filling factor and the strip spreading at the turn based on a DBR structure.
Теги: count rate current density diffusion coefficient meander topology quantum efficiency single-photon detectors x-ray amorphous materials квантовая эффективность коэффициент диффузии однофотонные детекторы плотность тока рентгеноаморфные материалы скорость счета топология меандра
Review of Superconducting Microstrip
Single-photon Detectors
V. A. Zheltikov, D. D. Platonov, S. Khydyrova, K. M. Moiseev, D. D. Vasilev
Bauman Moscow State Technical University, Moscow
This paper reviews the state-of-the-art of superconducting microstrip single-photon detectors and provides their comparison with the nanostrip detectors. The article describes the actual detection models required for understanding the SMSPD operating principles. We have analyzed the manufacturing materials of ultrathin films designed for SMSPD demonstrating the prospects for the use of X-ray amorphous materials with α-Mn structure. As a result of methods study to improve the detector specifications, we have provided the following recommendations: to reduce DCR and increase SDE, CR and active area, it is necessary to use the SMSPDs based on the X-ray amorphous materials with a low diffusion coefficient, a topology with a high filling factor and the strip spreading at the turn based on a DBR structure.
Keywords: single-photon detectors, X-ray amorphous materials, diffusion coefficient, quantum efficiency, count rate, current density, meander topology
Received on: 23.08.2022
Accepted on: 15.09.2022
INTRODUCTION
Due to their high-performance capabilities, the superconducting nanowire single-photon detectors (SNSPDs) are beneficial [1] for applications in various fields of science and technology, such as space communications, LIDAR systems, and quantum technologies. An obstacle to the mainstream use of SNSPDs at this stage of technological expansion has become the small size of the detecting element (the line width is about 100 nm, the meander area is about 10 × 10 μm2) and the low signal level (about 0.1 mV) [2]. This leads to the difficulties both in the manufacturing technology that is based on the expensive electron beam lithography equipment, and in the measurement procedures, where it is necessary to use the powerful signal amplifiers and lensed fiber for image focusing into a small detector working area. To manufacture SNSPDs, the crystalline materials (NbN, NbTiN, etc.) are mainly used [3] that often leads to a decreased yield ratio.
The main detector parameter is the quantum efficiency (QE, the ratio of the number of registered photons to the number of emitted photons) that is divided by the intrinsic absorption efficiency of the IDE film and the detection efficiency of the SDE system. The important parameters are the dark count rate (number of false detections per second, DCR), count rate (CR), and detection area. The devices are operated at low temperatures less than 10 K, since such temperatures are required for transition of the detecting element (ultrathin film) to the superconducting state. The infrastructure for the device operation includes the cryostats, the cost of which is increased by orders of magnitude when moving from the temperatures of 4.2 K to the temperatures of about 1 K. However, the decreased operating temperature leads to the increased QE and reduced DCR [4]. For this reason, most commercial detectors are used at the temperatures below 2.4 K [5]. To reduce the cost of devices, it is necessary to use materials with the critical temperatures above 4.2 K in their design.
DETECTION MODELS
For operation, the detector is cooled to the temperatures below 10 K and a bias current is passed to ensure that the photon energy is sufficient for local superconductivity destruction. When the strip goes into a resistive state, a potential pulse is observed on the oscilloscope that indicates the photon registration. Since 2001, the hot spot model has been widespread (proposed by G. N. Goltsman et al. [6]), according to which the detectors with a strip width greater than a certain critical dimension (about 200 nm) are not able to detect a photon. In 2017, D. Yu. Vodolazov proposed a vortex detection model [7], according to which the detection mechanism consists of three stages: at first, a photon is absorbed with a hot spot generation; then there is a penetration into the strip of a vortex-antivortex pair – dissipative motion of magnetic vortices to the strip edge under the influence of the Lorentz force; and finally, the strip section is heated to the normal state with the occurrence of a potential pulse (Fig. 1).
After proposal of this detection model, it became possible to use the detectors with a typical size of about 1 μm that simplified the device manufacturing technology. In 2018, the first superconducting microwire single photon detector (SMSPD) was produced [8]. Such devices have a high signal level of about 0.1 V that makes it possible to measure the detector response without the use of powerful amplifiers.
MATERIAL ANALYSIS FOR SMSPD
The vortex detection model leads to the requirements for the SMSPD materials: a low diffusion coefficient D < 0.5 cm2 / s and a small energy gap ∆ < 1 meV [2]. The economic considerations imply the need for a high critical temperature Tc > 4.2 K and an X-ray amorphous structure that is insensitive to the substrate defects and provides a higher yield ratio compared to the detectors based on crystalline materials [9].
At present, the crystalline materials (NbN, NbTiN, etc.) are used to produce the commercially available SNSPDs [3]. The research centers apply the X-ray amorphous materials, detectors, on the basis of which they demonstrate high efficiency (SDE 93% when using WSi [10]). The structure of the low-level cells of such films has been underinvestigated, however, it is indicated in [11] for the SNSPDs based on MoxSi(1‑x) that the formed film is amorphous and disordered, but has a structure similar in the short-range order to the crystal lattice of the A15, or β-W type. The A15 structure is also typical for the A3B crystalline compounds of X-ray amorphous materials used in SNSPDs: Mo3Ge, Mo3Re, Nb3Si [12, 13], as well as for tungsten and transition metal silicides and, presumably, for W3Si [14]. However, these materials in the amorphous state have either low Tc or high D and ∆ (see the table) and do not meet the requirements set for the detector production based on the microstrips. In 2020, the experts demonstrated the SMSPD detection based on X-ray amorphous NbRe [15] that had a structure of the α-Mn type [16] (stoichiometric compound A5B24). Such materials have been understudied. They are more promising for use in SMSPD, since they are superior to the materials with a β-W type structure in terms of the required parameters (see the table). For the SMSPD manufacture, it is proposed to use the previously unexplored Zr5Re24 detecting element material that has a high critical temperature of 4.9 K at a thickness of 8 nm, a small energy gap of 0.93 meV, and a diffusion coefficient of 0.46 cm2 / s.
In addition to the X-ray amorphous materials, the NbN crystalline material is also traditionally used to produce SMSPD [9, 25]. It has a higher critical temperature and, accordingly, a lower kinetic inductance, due to which the detectors have a high count rate (CR), but a small intrinsic efficiency IDE. In 2021, a manufacturing technology was proposed for SMSPD by irradiating crystalline NbN with the helium ions [25]. This solution led to the non-superconducting inclusions in the film being the centers of penetration of magnetic vortices that destroy superconductivity due to the photon ingress. Such films have a lower critical temperature than the non-irradiated films (6.4 and 7.1 K, respectively) leading to the decreased energy gap (i. e., a bias current sufficient for detection) and the increased QE. The given manufacturing method, as well as the use of a structure for reflecting photons of a given wavelength (distributed Bragg reflector, DBR), have made it possible to increase QE from 30 to 92.2% [25].
However, a detector manufactured according to this technology has a number of disadvantages. First of all, this is the use of a crystalline material that causes increased requirements for the film deposition and control process. Moreover, there is also the need for ion bombardment that is an expensive and comprehensive process operation. This leads to a low yield ratio and high cost of the device.
SMSPD PERFORMANCE UPGRADE METHODS
At present, the SNSPD detectors have the following record-setting parameters: SDE=99.5% [27], DCR=10 Hz [4], detection area: 0.07 mm2 [28], while the latest SMSPD demonstrate the following: SDE=96% [29], DCR=200 Hz [25] and detection area: 2.25×2.25 mm2 [30]. Such similar specifications are explained by the fact that the SMSPDs, although they appeared only in 2017, have already managed to run along the same development path as the SNSPDs since 2001 until the present time.
The current remaining challenge is a comprehensive increase in the SMSPD parameters [5]. Modern detectors most often demonstrate a record-setting value of only one parameter (for example, CR), while other specifications (QE, DCR) have low values.
The high detection efficiency is provided by the film-based detectors with good uniformity. This is explained by the fact that such films are capable of passing bias currents close to the critical current that lied at the bottom of the SMSPD detection principle. In this regard, the films made of amorphous materials such as MoSi and WSi have shown good results: the detectors based on these materials demonstrate high efficiency SDE = 93% [10], relatively low DCR = 1 kHz [31], and count rate CR = 10 MHz [32].
In addition to the materials that allow high bias currents to be passed, the topologies with a high filling factor, i. e. the ratio of the stripe width to the structure period (Fig. 2a), are used to increase QE. In [29] published in 2022, the researchers proposed a new meander design in the form of a candelabra-style with a filling factor of 0.91 in the active region. This solution makes it possible to increase the probability of a photon hitting the superconducting strip rather than the gap region. Thus, the detector efficiency using the DBR structure is 96% [29].
Another way to increase the detection efficiency is to increase the probability of a photon absorption by a meander. For this purpose, a quarter-wave resonator (Fig. 2b) or a DBR structure (Fig. 2c) is added to the structure when producing both types of detectors. A quarter-wave resonator is an optical cavity made of a dielectric material (usually SiO due to its high transparency for IR radiation) and a reflective coating made of gold or titanium [33]. The radiation that has not been absorbed by the detector is reflected from the mirror and begins to resonate in the optical cavity, thus increasing the photon absorption probability upon its entering the meander gap region. The DBR structure consists of several layers of materials with various refractive indices, such as SiO2 and Ta2O5. The thickness is selected depending on the wavelength at which the detector will be operated, and the number of layers affects the reflection probability. Moreover, the structure with such a reflector increases the probability of photon absorption by the meander that leads to the increased SDE [25]. Sometimes it is possible to use the structures combining both elements. Although their manufacturing method is more comprehensive, however, in this case, the photon absorption probability is close to 100% [34].
In order to reduce DCR, the scientific teams propose new detector topologies that increase the current density uniformity during the device operation period. This is due to the fact that at high bias currents required for the SMSPD operation, the number of fluctuations that can destroy superconductivity in the absence of a photon (that is, increase DCR) is increased. The highest current density is observed at the meander turns due to the fact that it is energetically more profitable for the electrons to negotiate the turn along the shortest path that is the inner radius (the current is passed along the path of least resistance).
Many scientific teams use a spiral (Fig. 3a) in their work as a detector topology, since the smooth constant spiral involution provides a uniform current distribution over the strip [25, 26]. In 2021, there was a proposal to use thickenings at the meander turns (Fig. 3b) that reduced the current distribution unevenness by 20% leading to the decreased DCR [33, 35]. The main disadvantage of the described technology is the labor-intensive manufacturing process, since thickening requires an additional lithography process, as a result of which the strip edge is rough and reduces QE. In 2022, a «single-layer» topology with the turn broadening was proposed (Fig. 3c), corresponding to the standard SNSPD manufacturing process [36].
Figure 4 compares the main topologies of the SNSPD meander in terms of the uneven current density distribution [36]. Compared to the standard one (meander), the new geometries (spiral, meander with the turn thickening or broadening) increase the current uniformity at the turns from 67 to 98%, while reducing DCR and increasing the detector efficiency.
The detector’s count rate is the number of photons per second to be recorded by the detector. This parameter is affected by the detector’s recovery time, that is, the time during which it changes from the normal to the superconducting state and is ready for the repeated photon detection. To reduce it, the materials with a low kinetic inductance and a high critical temperature are used, giving preference to the crystalline films (NbN, NbTiN) [3]. However, the X-ray amorphous films with a low diffusion coefficient (for example, MoSi at D = 0.47 cm2 / s) show good results with CR=10 MHz [32]. We assume that the high count rate is related to a short recovery time due to the weak electron-electron interaction in the film and the rapid heat removal to the substrate. Moreover, to reduce the kinetic inductance, it is necessary to use the topologies with a minimum meander length. The best solution for the increased CR is a microbridge (Fig. 5a) [37, 38].
The main advantage of SMSPD compared to SNSPD is the possible increased of the detector’s active area without increasing the meander length. A large active area is required for such tasks as the search for dark matter or the detection of neutrons and macromolecules [30]. Usually, the large-area of the produced SNSPD leads to the great increase in the meander length entailing an increase in the kinetic inductance and a decrease in the detector’s CR. Therefore, the standard SNSPD area is about 10×10 µm2, the maximum area is about 265×265 µm2 (Fig. 5b) at CR=10 MHz [28]. With a standard active area, the detector needs precise alignment with the single-mode optical fiber. The SMSPD, due to a stripe width of about 2 µm, has an area of about 600 × 600 µm2 (Fig. 5c) at CR=8 MHz [39], comparable to SNSPD. This result eliminates the need for precision detector-fiber alignment systems.
CONCLUSION
At present, the superconducting single-photon detectors SNSPD are the most efficient photon detectors. Since 2001 they have been produced using the crystalline materials and meanders with a strip width of about 100 nm. However, after the discovery of the vortex detection model in 2017, it became possible to switch to the SMSPD detectors made of X-ray amorphous materials and with a strip width of about 1 μm that had lower requirements for film deposition and lithography. Being a new stage in the SNSPD development, they lead to a decrease in the cost of the device with the similar (SDE, DCR) and better (CR, detection area) specifications.
According to the results of the literature review, it is proposed to manufacture the detecting element of SMSPD from Zr5Re24 material that has not been previously studied for SNSPD, and has an X-ray amorphous structure of the α-Mn type. To improve QE and reduce DCR, it is recommended to use the topology with the turn broadening, high filling factor and DBR structure. To increase CR, it is possible to manufacture the detector in the form of a microbridge, reducing the kinetic inductance to a minimum level.
AUTHORS
V. A. Zheltikov, 2nd year Master’s student, zheltikov.vladimir@yandex.ru, Bauman Moscow State Technical University, Moscow.
ORCID 0000–0001–7099–1039
D. D. Platonov, 4th year Bachelor’s student, Bauman Moscow State Technical University, Moscow.
ORCID 0000–0003–0246–4290
S. Hydyrova, 2nd year Postgraduate’s student, Bauman Moscow State Technical University, Moscow.
ORCID: 0000–0002–5510–0899
K. M. Moiseev, Cand. of Tech. Sc., Docent, Bauman Moscow State Technical University, Moscow.
ORCID 0000–0002–8753–7737
D. D. Vasilev, Cand. of Tech. Sc., Docent, Bauman Moscow State Technical University, Moscow.
ORCID 0000–0003–2147–4216
Single-photon Detectors
V. A. Zheltikov, D. D. Platonov, S. Khydyrova, K. M. Moiseev, D. D. Vasilev
Bauman Moscow State Technical University, Moscow
This paper reviews the state-of-the-art of superconducting microstrip single-photon detectors and provides their comparison with the nanostrip detectors. The article describes the actual detection models required for understanding the SMSPD operating principles. We have analyzed the manufacturing materials of ultrathin films designed for SMSPD demonstrating the prospects for the use of X-ray amorphous materials with α-Mn structure. As a result of methods study to improve the detector specifications, we have provided the following recommendations: to reduce DCR and increase SDE, CR and active area, it is necessary to use the SMSPDs based on the X-ray amorphous materials with a low diffusion coefficient, a topology with a high filling factor and the strip spreading at the turn based on a DBR structure.
Keywords: single-photon detectors, X-ray amorphous materials, diffusion coefficient, quantum efficiency, count rate, current density, meander topology
Received on: 23.08.2022
Accepted on: 15.09.2022
INTRODUCTION
Due to their high-performance capabilities, the superconducting nanowire single-photon detectors (SNSPDs) are beneficial [1] for applications in various fields of science and technology, such as space communications, LIDAR systems, and quantum technologies. An obstacle to the mainstream use of SNSPDs at this stage of technological expansion has become the small size of the detecting element (the line width is about 100 nm, the meander area is about 10 × 10 μm2) and the low signal level (about 0.1 mV) [2]. This leads to the difficulties both in the manufacturing technology that is based on the expensive electron beam lithography equipment, and in the measurement procedures, where it is necessary to use the powerful signal amplifiers and lensed fiber for image focusing into a small detector working area. To manufacture SNSPDs, the crystalline materials (NbN, NbTiN, etc.) are mainly used [3] that often leads to a decreased yield ratio.
The main detector parameter is the quantum efficiency (QE, the ratio of the number of registered photons to the number of emitted photons) that is divided by the intrinsic absorption efficiency of the IDE film and the detection efficiency of the SDE system. The important parameters are the dark count rate (number of false detections per second, DCR), count rate (CR), and detection area. The devices are operated at low temperatures less than 10 K, since such temperatures are required for transition of the detecting element (ultrathin film) to the superconducting state. The infrastructure for the device operation includes the cryostats, the cost of which is increased by orders of magnitude when moving from the temperatures of 4.2 K to the temperatures of about 1 K. However, the decreased operating temperature leads to the increased QE and reduced DCR [4]. For this reason, most commercial detectors are used at the temperatures below 2.4 K [5]. To reduce the cost of devices, it is necessary to use materials with the critical temperatures above 4.2 K in their design.
DETECTION MODELS
For operation, the detector is cooled to the temperatures below 10 K and a bias current is passed to ensure that the photon energy is sufficient for local superconductivity destruction. When the strip goes into a resistive state, a potential pulse is observed on the oscilloscope that indicates the photon registration. Since 2001, the hot spot model has been widespread (proposed by G. N. Goltsman et al. [6]), according to which the detectors with a strip width greater than a certain critical dimension (about 200 nm) are not able to detect a photon. In 2017, D. Yu. Vodolazov proposed a vortex detection model [7], according to which the detection mechanism consists of three stages: at first, a photon is absorbed with a hot spot generation; then there is a penetration into the strip of a vortex-antivortex pair – dissipative motion of magnetic vortices to the strip edge under the influence of the Lorentz force; and finally, the strip section is heated to the normal state with the occurrence of a potential pulse (Fig. 1).
After proposal of this detection model, it became possible to use the detectors with a typical size of about 1 μm that simplified the device manufacturing technology. In 2018, the first superconducting microwire single photon detector (SMSPD) was produced [8]. Such devices have a high signal level of about 0.1 V that makes it possible to measure the detector response without the use of powerful amplifiers.
MATERIAL ANALYSIS FOR SMSPD
The vortex detection model leads to the requirements for the SMSPD materials: a low diffusion coefficient D < 0.5 cm2 / s and a small energy gap ∆ < 1 meV [2]. The economic considerations imply the need for a high critical temperature Tc > 4.2 K and an X-ray amorphous structure that is insensitive to the substrate defects and provides a higher yield ratio compared to the detectors based on crystalline materials [9].
At present, the crystalline materials (NbN, NbTiN, etc.) are used to produce the commercially available SNSPDs [3]. The research centers apply the X-ray amorphous materials, detectors, on the basis of which they demonstrate high efficiency (SDE 93% when using WSi [10]). The structure of the low-level cells of such films has been underinvestigated, however, it is indicated in [11] for the SNSPDs based on MoxSi(1‑x) that the formed film is amorphous and disordered, but has a structure similar in the short-range order to the crystal lattice of the A15, or β-W type. The A15 structure is also typical for the A3B crystalline compounds of X-ray amorphous materials used in SNSPDs: Mo3Ge, Mo3Re, Nb3Si [12, 13], as well as for tungsten and transition metal silicides and, presumably, for W3Si [14]. However, these materials in the amorphous state have either low Tc or high D and ∆ (see the table) and do not meet the requirements set for the detector production based on the microstrips. In 2020, the experts demonstrated the SMSPD detection based on X-ray amorphous NbRe [15] that had a structure of the α-Mn type [16] (stoichiometric compound A5B24). Such materials have been understudied. They are more promising for use in SMSPD, since they are superior to the materials with a β-W type structure in terms of the required parameters (see the table). For the SMSPD manufacture, it is proposed to use the previously unexplored Zr5Re24 detecting element material that has a high critical temperature of 4.9 K at a thickness of 8 nm, a small energy gap of 0.93 meV, and a diffusion coefficient of 0.46 cm2 / s.
In addition to the X-ray amorphous materials, the NbN crystalline material is also traditionally used to produce SMSPD [9, 25]. It has a higher critical temperature and, accordingly, a lower kinetic inductance, due to which the detectors have a high count rate (CR), but a small intrinsic efficiency IDE. In 2021, a manufacturing technology was proposed for SMSPD by irradiating crystalline NbN with the helium ions [25]. This solution led to the non-superconducting inclusions in the film being the centers of penetration of magnetic vortices that destroy superconductivity due to the photon ingress. Such films have a lower critical temperature than the non-irradiated films (6.4 and 7.1 K, respectively) leading to the decreased energy gap (i. e., a bias current sufficient for detection) and the increased QE. The given manufacturing method, as well as the use of a structure for reflecting photons of a given wavelength (distributed Bragg reflector, DBR), have made it possible to increase QE from 30 to 92.2% [25].
However, a detector manufactured according to this technology has a number of disadvantages. First of all, this is the use of a crystalline material that causes increased requirements for the film deposition and control process. Moreover, there is also the need for ion bombardment that is an expensive and comprehensive process operation. This leads to a low yield ratio and high cost of the device.
SMSPD PERFORMANCE UPGRADE METHODS
At present, the SNSPD detectors have the following record-setting parameters: SDE=99.5% [27], DCR=10 Hz [4], detection area: 0.07 mm2 [28], while the latest SMSPD demonstrate the following: SDE=96% [29], DCR=200 Hz [25] and detection area: 2.25×2.25 mm2 [30]. Such similar specifications are explained by the fact that the SMSPDs, although they appeared only in 2017, have already managed to run along the same development path as the SNSPDs since 2001 until the present time.
The current remaining challenge is a comprehensive increase in the SMSPD parameters [5]. Modern detectors most often demonstrate a record-setting value of only one parameter (for example, CR), while other specifications (QE, DCR) have low values.
The high detection efficiency is provided by the film-based detectors with good uniformity. This is explained by the fact that such films are capable of passing bias currents close to the critical current that lied at the bottom of the SMSPD detection principle. In this regard, the films made of amorphous materials such as MoSi and WSi have shown good results: the detectors based on these materials demonstrate high efficiency SDE = 93% [10], relatively low DCR = 1 kHz [31], and count rate CR = 10 MHz [32].
In addition to the materials that allow high bias currents to be passed, the topologies with a high filling factor, i. e. the ratio of the stripe width to the structure period (Fig. 2a), are used to increase QE. In [29] published in 2022, the researchers proposed a new meander design in the form of a candelabra-style with a filling factor of 0.91 in the active region. This solution makes it possible to increase the probability of a photon hitting the superconducting strip rather than the gap region. Thus, the detector efficiency using the DBR structure is 96% [29].
Another way to increase the detection efficiency is to increase the probability of a photon absorption by a meander. For this purpose, a quarter-wave resonator (Fig. 2b) or a DBR structure (Fig. 2c) is added to the structure when producing both types of detectors. A quarter-wave resonator is an optical cavity made of a dielectric material (usually SiO due to its high transparency for IR radiation) and a reflective coating made of gold or titanium [33]. The radiation that has not been absorbed by the detector is reflected from the mirror and begins to resonate in the optical cavity, thus increasing the photon absorption probability upon its entering the meander gap region. The DBR structure consists of several layers of materials with various refractive indices, such as SiO2 and Ta2O5. The thickness is selected depending on the wavelength at which the detector will be operated, and the number of layers affects the reflection probability. Moreover, the structure with such a reflector increases the probability of photon absorption by the meander that leads to the increased SDE [25]. Sometimes it is possible to use the structures combining both elements. Although their manufacturing method is more comprehensive, however, in this case, the photon absorption probability is close to 100% [34].
In order to reduce DCR, the scientific teams propose new detector topologies that increase the current density uniformity during the device operation period. This is due to the fact that at high bias currents required for the SMSPD operation, the number of fluctuations that can destroy superconductivity in the absence of a photon (that is, increase DCR) is increased. The highest current density is observed at the meander turns due to the fact that it is energetically more profitable for the electrons to negotiate the turn along the shortest path that is the inner radius (the current is passed along the path of least resistance).
Many scientific teams use a spiral (Fig. 3a) in their work as a detector topology, since the smooth constant spiral involution provides a uniform current distribution over the strip [25, 26]. In 2021, there was a proposal to use thickenings at the meander turns (Fig. 3b) that reduced the current distribution unevenness by 20% leading to the decreased DCR [33, 35]. The main disadvantage of the described technology is the labor-intensive manufacturing process, since thickening requires an additional lithography process, as a result of which the strip edge is rough and reduces QE. In 2022, a «single-layer» topology with the turn broadening was proposed (Fig. 3c), corresponding to the standard SNSPD manufacturing process [36].
Figure 4 compares the main topologies of the SNSPD meander in terms of the uneven current density distribution [36]. Compared to the standard one (meander), the new geometries (spiral, meander with the turn thickening or broadening) increase the current uniformity at the turns from 67 to 98%, while reducing DCR and increasing the detector efficiency.
The detector’s count rate is the number of photons per second to be recorded by the detector. This parameter is affected by the detector’s recovery time, that is, the time during which it changes from the normal to the superconducting state and is ready for the repeated photon detection. To reduce it, the materials with a low kinetic inductance and a high critical temperature are used, giving preference to the crystalline films (NbN, NbTiN) [3]. However, the X-ray amorphous films with a low diffusion coefficient (for example, MoSi at D = 0.47 cm2 / s) show good results with CR=10 MHz [32]. We assume that the high count rate is related to a short recovery time due to the weak electron-electron interaction in the film and the rapid heat removal to the substrate. Moreover, to reduce the kinetic inductance, it is necessary to use the topologies with a minimum meander length. The best solution for the increased CR is a microbridge (Fig. 5a) [37, 38].
The main advantage of SMSPD compared to SNSPD is the possible increased of the detector’s active area without increasing the meander length. A large active area is required for such tasks as the search for dark matter or the detection of neutrons and macromolecules [30]. Usually, the large-area of the produced SNSPD leads to the great increase in the meander length entailing an increase in the kinetic inductance and a decrease in the detector’s CR. Therefore, the standard SNSPD area is about 10×10 µm2, the maximum area is about 265×265 µm2 (Fig. 5b) at CR=10 MHz [28]. With a standard active area, the detector needs precise alignment with the single-mode optical fiber. The SMSPD, due to a stripe width of about 2 µm, has an area of about 600 × 600 µm2 (Fig. 5c) at CR=8 MHz [39], comparable to SNSPD. This result eliminates the need for precision detector-fiber alignment systems.
CONCLUSION
At present, the superconducting single-photon detectors SNSPD are the most efficient photon detectors. Since 2001 they have been produced using the crystalline materials and meanders with a strip width of about 100 nm. However, after the discovery of the vortex detection model in 2017, it became possible to switch to the SMSPD detectors made of X-ray amorphous materials and with a strip width of about 1 μm that had lower requirements for film deposition and lithography. Being a new stage in the SNSPD development, they lead to a decrease in the cost of the device with the similar (SDE, DCR) and better (CR, detection area) specifications.
According to the results of the literature review, it is proposed to manufacture the detecting element of SMSPD from Zr5Re24 material that has not been previously studied for SNSPD, and has an X-ray amorphous structure of the α-Mn type. To improve QE and reduce DCR, it is recommended to use the topology with the turn broadening, high filling factor and DBR structure. To increase CR, it is possible to manufacture the detector in the form of a microbridge, reducing the kinetic inductance to a minimum level.
AUTHORS
V. A. Zheltikov, 2nd year Master’s student, zheltikov.vladimir@yandex.ru, Bauman Moscow State Technical University, Moscow.
ORCID 0000–0001–7099–1039
D. D. Platonov, 4th year Bachelor’s student, Bauman Moscow State Technical University, Moscow.
ORCID 0000–0003–0246–4290
S. Hydyrova, 2nd year Postgraduate’s student, Bauman Moscow State Technical University, Moscow.
ORCID: 0000–0002–5510–0899
K. M. Moiseev, Cand. of Tech. Sc., Docent, Bauman Moscow State Technical University, Moscow.
ORCID 0000–0002–8753–7737
D. D. Vasilev, Cand. of Tech. Sc., Docent, Bauman Moscow State Technical University, Moscow.
ORCID 0000–0003–2147–4216
Readers feedback
